Noise-Powered Disentangled Representation for Unsupervised Speckle Reduction of Optical Coherence Tomography Images

Unsupervised Domain Adaptation for EM Image Denoising With Invertible Networks

Electron microscopy (EM) image denoising is critical for visualization and subsequent analysis. Despite the remarkable achievements of deep learning-based non-blind denoising methods, their performance drops significantly when domain shifts exist between the training and testing data. To address this issue, unpaired blind denoising methods have been proposed. However, these methods heavily rely on image-to-image translation and neglect the inherent characteristics of EM images, limiting their overall denoising performance. In this paper, we propose the first unsupervised domain adaptive EM image denoising method, which is grounded in the observation that EM images from similar samples share common content characteristics. Specifically, we first disentangle the content representations and the noise components from noisy images and establish a shared domain-agnostic content space via domain alignment to bridge the synthetic images (source domain) and the real images (target domain). To ensure precise domain alignment, we further incorporate domain regularization by enforcing that: the pseudo-noisy images, reconstructed using both content representations and noise components, accurately capture the characteristics of the noisy images from which the noise components originate, all while maintaining semantic consistency with the noisy images from which the content representations originate. To guarantee lossless representation decomposition and image reconstruction, we introduce disentanglement-reconstruction invertible networks. Finally, the reconstructed pseudo-noisy images, paired with their corresponding clean counterparts, serve as valuable training data for the denoising network. Extensive experiments on synthetic and real EM datasets demonstrate the superiority of our method in terms of image restoration quality and downstream neuron segmentation accuracy. Our code is publicly available at https://github.com/sydeng99/DADn.

Noise-imitation learning: unpaired speckle noise reduction for optical coherence tomography

Objective.Optical coherence tomography (OCT) is widely used in clinical practice for its non-invasive, high-resolution imaging capabilities. However, speckle noise inherent to its low coherence principle can degrade image quality and compromise diagnostic accuracy. While deep learning methods have shown promise in reducing speckle noise, obtaining well-registered image pairs remains challenging, leading to the development of unpaired methods. Despite their potential, existing unpaired methods suffer from redundancy in network structures or interaction mechanisms. Therefore, a more streamlined method for unpaired OCT denoising is essential.Approach.In this work, we propose a novel unpaired method for OCT image denoising, referred to as noise-imitation learning (NIL). NIL comprises three primary modules: the noise extraction module, which extracts noise features by denoising noisy images; the noise imitation module, which synthesizes noisy images and generates fake clean images; and the adversarial learning module, which differentiates between real and fake clean images through adversarial training. The complexity of NIL is significantly lower than that of previous unpaired methods, utilizing only one generator and one discriminator for training.Main results.By efficiently fusing unpaired images and employing adversarial training, NIL can extract more speckle noise information to enhance denoising performance. Building on NIL, we propose an OCT image denoising pipeline, NIL-NAFNet. This pipeline achieved PSNR, SSIM, and RMSE values of 31.27 dB, 0.865, and 7.00, respectively, on the PKU37 dataset. Extensive experiments suggest that our method outperforms state-of-the-art unpaired methods both qualitatively and quantitatively.Significance.These findings indicate that the proposed NIL is a simple yet effective method for unpaired OCT speckle noise reduction. The OCT denoising pipeline based on NIL demonstrates exceptional performance and efficiency. By addressing speckle noise without requiring well-registered image pairs, this method can enhance image quality and diagnostic accuracy in clinical practice.

Tensor Ring Decomposition Guided Dictionary Learning for OCT Image Denoising

Optical coherence tomography (OCT) is a non-invasive and effective tool for the imaging of retinal tissue. However, the heavy speckle noise, resulting from multiple scattering of the light waves, obscures important morphological structures and impairs the clinical diagnosis of ocular diseases. In this paper, we propose a novel and powerful model known as tensor ring decomposition-guided dictionary learning (TRGDL) for OCT image denoising, which can simultaneously utilize two useful complementary priors, i.e., three-dimensional low-rank and sparsity priors, under a unified framework. Specifically, to effectively use the strong correlation between nearby OCT frames, we construct the OCT group tensors by extracting cubic patches from OCT images and clustering similar patches. Then, since each created OCT group tensor has a low-rank structure, to exploit spatial, non-local, and its temporal correlations in a balanced way, we enforce the TR decomposition model on each OCT group tensor. Next, to use the beneficial three-dimensional inter-group sparsity, we learn shared dictionaries in both spatial and temporal dimensions from all of the stacked OCT group tensors. Furthermore, we develop an effective algorithm to solve the resulting optimization problem by using two efficient optimization approaches, including proximal alternating minimization and the alternative direction method of multipliers. Finally, extensive experiments on OCT datasets from various imaging devices are conducted to prove the generality and usefulness of the proposed TRGDL model. Experimental simulation results show that the suggested TRGDL model outperforms state-of-the-art approaches for OCT image denoising both qualitatively and quantitatively.

Total variation regularized tensor ring decomposition for OCT image denoising and super-resolution

This paper suggests a novel hybrid tensor-ring (TR) decomposition and first-order tensor-based total variation (FOTTV) model, known as the TRFOTTV model, for super-resolution and noise suppression of optical coherence tomography (OCT) images. OCT imaging faces two fundamental problems undermining correct OCT-based diagnosis: significant noise levels and low sampling rates to speed up the capturing process. Inspired by the effectiveness of TR decomposition in analyzing complicated data structures, we suggest the TRFOTTV model for noise suppression and super-resolution of OCT images. Initially, we extract the nonlocal 3D patches from OCT data and group them to create a third-order low-rank tensor. Subsequently, using TR decomposition, we extract the correlations among all modes of the grouped OCT tensor. Finally, FOTTV is integrated into the TR model to enhance spatial smoothness in OCT images and conserve layer structures more effectively. The proximal alternating minimization and alternative direction method of multipliers are applied to solve the obtained optimization problem. The effectiveness of the suggested method is verified by four OCT datasets, demonstrating superior visual and numerical outcomes compared to state-of-the-art procedures.

Parallel processing model for low-dose computed tomography image denoising

Low-dose computed tomography (LDCT) has gained increasing attention owing to its crucial role in reducing radiation exposure in patients. However, LDCT-reconstructed images often suffer from significant noise and artifacts, negatively impacting the radiologists' ability to accurately diagnose. To address this issue, many studies have focused on denoising LDCT images using deep learning (DL) methods. However, these DL-based denoising methods have been hindered by the highly variable feature distribution of LDCT data from different imaging sources, which adversely affects the performance of current denoising models. In this study, we propose a parallel processing model, the multi-encoder deep feature transformation network (MDFTN), which is designed to enhance the performance of LDCT imaging for multisource data. Unlike traditional network structures, which rely on continual learning to process multitask data, the approach can simultaneously handle LDCT images within a unified framework from various imaging sources. The proposed MDFTN consists of multiple encoders and decoders along with a deep feature transformation module (DFTM). During forward propagation in network training, each encoder extracts diverse features from its respective data source in parallel and the DFTM compresses these features into a shared feature space. Subsequently, each decoder performs an inverse operation for multisource loss estimation. Through collaborative training, the proposed MDFTN leverages the complementary advantages of multisource data distribution to enhance its adaptability and generalization. Numerous experiments were conducted on two public datasets and one local dataset, which demonstrated that the proposed network model can simultaneously process multisource data while effectively suppressing noise and preserving fine structures. The source code is available at https://github.com/123456789ey/MDFTN .

Toward Ground-Truth Optical Coherence Tomography via Three-Dimensional Unsupervised Deep Learning Processing and Data

Optical coherence tomography (OCT) can perform non-invasive high-resolution three-dimensional (3D) imaging and has been widely used in biomedical fields, while it is inevitably affected by coherence speckle noise which degrades OCT imaging performance and restricts its applications. Here we present a novel speckle-free OCT imaging strategy, named toward-ground-truth OCT ( t GT-OCT), that utilizes unsupervised 3D deep-learning processing and leverages OCT 3D imaging features to achieve speckle-free OCT imaging. Specifically, our proposed t GT-OCT utilizes an unsupervised 3D-convolution deep-learning network trained using random 3D volumetric data to distinguish and separate speckle from real structures in 3D imaging volumetric space; moreover, t GT-OCT effectively further reduces speckle noise and reveals structures that would otherwise be obscured by speckle noise while preserving spatial resolution. Results derived from different samples demonstrated the high-quality speckle-free 3D imaging performance of t GT-OCT and its advancement beyond the previous state-of-the-art. The code is available online: https://github.com/Voluntino/tGT-OCT.

Enhancement of OCTen faceimages by unsupervised deep learning

Objective. The quality of optical coherence tomography (OCT)en faceimages is crucial for clinical visualization of early disease. As a three dimensional and coherent imaging, defocus and speckle noise are inevitable, which seriously affect evaluation of microstructure of bio-samples in OCT images. The deep learning has demonstrated great potential in OCT refocusing and denoising, but it is limited by the difficulty of sufficient paired training data. This work aims to develop an unsupervised method to enhance the quality of OCTen faceimages.Approach. We proposed an unsupervised deep learning-based pipeline. The unregistered defocused conventional OCT images and focused speckle-free OCT images were collected by a home-made speckle modulating OCT system to construct the dataset. The image enhancement model was trained with the cycle training strategy. Finally, the speckle noise and defocus were both effectively improved.Main results. The experimental results on complex bio-samples indicated that the proposed method is effective and generalized in enhancing the quality of OCTen faceimages.Significance. The proposed unsupervised deep learning method helps to reduce the complexity of data construction, which is conducive to practical applications in OCT bio-sample imaging.

Masked autoencoders with generalizable self-distillation for skin lesion segmentation

In the field of skin lesion image segmentation, accurate identification and partitioning of diseased regions is of vital importance for in-depth analysis of skin cancer. Self-supervised learning, i.e., MAE, has emerged as a potent force in the medical imaging domain, which autonomously learns and extracts latent features from unlabeled data, thereby yielding pre-trained models that greatly assist downstream tasks. To encourage pre-trained models to more comprehensively learn the global structural and local detail information inherent in dermoscopy images, we introduce a Teacher-Student architecture, named TEDMAE, by incorporating a self-distillation mechanism, it learns holistic image feature information to improve the generalizable global knowledge learning of the student MAE model. To make the image features learned by the model suitable for unknown test images, two optimization strategies are, Exterior Conversion Augmentation (EC) utilizes random convolutional kernels and linear interpolation to effectively transform the input image into one with the same shape but altered intensities and textures, while Dynamic Feature Generation (DF) employs a nonlinear attention mechanism for feature merging, enhancing the expressive power of the features, are proposed to enhance the generalizability of global features learned by the teacher model, thereby improving the overall generalization capability of the pre-trained models. Experimental results from the three public skin disease datasets, ISIC2019, ISIC2017, and PH2 indicate that our proposed TEDMAE method outperforms several similar approaches. Specifically, TEDMAE demonstrated optimal segmentation and generalization performance on the ISIC2017 and PH2 datasets, with Dice scores reaching 82.1% and 91.2%, respectively. The best Jaccard values were 72.6% and 84.5%, while the optimal HD95% values were 13.0% and 8.9%, respectively.

Denoising of Optical Coherence Tomography Images in Ophthalmology Using Deep Learning: A Systematic Review

Imaging from optical coherence tomography (OCT) is widely used for detecting retinal diseases, localization of intra-retinal boundaries, etc. It is, however, degraded by speckle noise. Deep learning models can aid with denoising, allowing clinicians to clearly diagnose retinal diseases. Deep learning models can be considered as an end-to-end framework. We selected denoising studies that used deep learning models with retinal OCT imagery. Each study was quality-assessed through image quality metrics (including the peak signal-to-noise ratio-PSNR, contrast-to-noise ratio-CNR, and structural similarity index metric-SSIM). Meta-analysis could not be performed due to heterogeneity in the methods of the studies and measurements of their performance. Multiple databases (including Medline via PubMed, Google Scholar, Scopus, Embase) and a repository (ArXiv) were screened for publications published after 2010, without any limitation on language. From the 95 potential studies identified, a total of 41 were evaluated thoroughly. Fifty-four of these studies were excluded after full text assessment depending on whether deep learning (DL) was utilized or the dataset and results were not effectively explained. Numerous types of OCT images are mentioned in this review consisting of public retinal image datasets utilized purposefully for denoising OCT images (n = 37) and the Optic Nerve Head (ONH) (n = 4). A wide range of image quality metrics was used; PSNR and SNR that ranged between 8 and 156 dB. The minority of studies (n = 8) showed a low risk of bias in all domains. Studies utilizing ONH images produced either a PSNR or SNR value varying from 8.1 to 25.7 dB, and that of public retinal datasets was 26.4 to 158.6 dB. Further analysis on denoising models was not possible due to discrepancies in reporting that did not allow useful pooling. An increasing number of studies have investigated denoising retinal OCT images using deep learning, with a range of architectures being implemented. The reported increase in image quality metrics seems promising, while study and reporting quality are currently low.

Unsupervised OCT image despeckling with ground-truth- and repeated-scanning-free features

Optical coherence tomography (OCT) can resolve biological three-dimensional tissue structures, but it is inevitably plagued by speckle noise that degrades image quality and obscures biological structure. Recently unsupervised deep learning methods are becoming more popular in OCT despeckling but they still have to use unpaired noisy-clean images or paired noisy-noisy images. To address the above problem, we propose what we believe to be a novel unsupervised deep learning method for OCT despeckling, termed Double-free Net, which eliminates the need for ground truth data and repeated scanning by sub-sampling noisy images and synthesizing noisier images. In comparison to existing unsupervised methods, Double-free Net obtains superior denoising performance when trained on datasets comprising retinal and human tissue images without clean images. The efficacy of Double-free Net in denoising holds significant promise for diagnostic applications in retinal pathologies and enhances the accuracy of retinal layer segmentation. Results demonstrate that Double-free Net outperforms state-of-the-art methods and exhibits strong convenience and adaptability across different OCT images.

Self-supervised Self2Self denoising strategy for OCT speckle reduction with a single noisy image

Optical coherence tomography (OCT) inevitably suffers from the influence of speckles originating from multiple scattered photons owing to its low-coherence interferometry property. Although various deep learning schemes have been proposed for OCT despeckling, they typically suffer from the requirement for ground-truth images, which are difficult to collect in clinical practice. To alleviate the influences of speckles without requiring ground-truth images, this paper presents a self-supervised deep learning scheme, namely, Self2Self strategy (S2Snet), for OCT despeckling using a single noisy image. Specifically, in this study, the main deep learning architecture is the Self2Self network, with its partial convolution being updated with a gated convolution layer. Specifically, both the input images and their Bernoulli sampling instances are adopted as network input first, and then, a devised loss function is integrated into the network to remove the background noise. Finally, the denoised output is estimated using the average of multiple predicted outputs. Experiments with various OCT datasets are conducted to verify the effectiveness of the proposed S2Snet scheme. Results compared with those of the existing methods demonstrate that S2Snet not only outperforms those existing self-supervised deep learning methods but also achieves better performances than those non-deep learning ones in different cases. Specifically, S2Snet achieves an improvement of 3.41% and 2.37% for PSNR and SSIM, respectively, as compared to the original Self2Self network, while such improvements become 19.9% and 22.7% as compared with the well-known non-deep learning NWSR method.

Pseudoaveraging for denoising of OCT angiography: a deep learning approach for image quality enhancement in healthy and diabetic eyes

BACKGROUND

This study aimed to develop a deep learning (DL) algorithm that enhances the quality of a single-frame enface OCTA scan to make it comparable to 4-frame averaged scan without the need for the repeated acquisitions required for averaging.

METHODS

Each of the healthy eyes and eyes from diabetic subjects that were prospectively enrolled in this cross-sectional study underwent four repeated 6 × 6 mm macular scans (PLEX Elite 9000 SS-OCT), and the repeated scans of each eye were co-registered to produce 4-frame averages. This prospective dataset of original (single-frame) enface scans and their corresponding averaged scans was divided into a training dataset and a validation dataset. In the training dataset, a DL algorithm (named pseudoaveraging) was trained using original scans as input and 4-frame averages as target. In the validation dataset, the pseudoaveraging algorithm was applied to single-frame scans to produce pseudoaveraged scans, and the single-frame and its corresponding averaged and pseudoaveraged scans were all qualitatively compared. In a separate retrospectively collected dataset of single-frame scans from eyes of diabetic subjects, the DL algorithm was applied, and the produced pseudoaveraged scan was qualitatively compared against its corresponding original.

RESULTS

This study included 39 eyes that comprised the prospective dataset (split into 5 eyes for training and 34 eyes for validating the DL algorithm), and 105 eyes that comprised the retrospective test dataset. Of the total 144 study eyes, 58% had any level of diabetic retinopathy (with and without diabetic macular edema), and the rest were from healthy eyes or eyes of diabetic subjects but without diabetic retinopathy and without macular edema. Grading results in the validation dataset showed that the pseudoaveraged enface scan ranked best in overall scan quality, background noise reduction, and visibility of microaneurysms (p < 0.05). Averaged scan ranked best for motion artifact reduction (p < 0.05). Grading results in the test dataset showed that pseudoaveraging resulted in enhanced small vessels, reduction of background noise, and motion artifact in 100%, 82%, and 98% of scans, respectively. Rates of false-positive/-negative perfusion were zero.

CONCLUSION

Pseudoaveraging is a feasible DL approach to more efficiently improve enface OCTA scan quality without introducing notable image artifacts.

Frequency-aware optical coherence tomography image super-resolution via conditional generative adversarial neural network

Optical coherence tomography (OCT) has stimulated a wide range of medical image-based diagnosis and treatment in fields such as cardiology and ophthalmology. Such applications can be further facilitated by deep learning-based super-resolution technology, which improves the capability of resolving morphological structures. However, existing deep learning-based method only focuses on spatial distribution and disregards frequency fidelity in image reconstruction, leading to a frequency bias. To overcome this limitation, we propose a frequency-aware super-resolution framework that integrates three critical frequency-based modules (i.e., frequency transformation, frequency skip connection, and frequency alignment) and frequency-based loss function into a conditional generative adversarial network (cGAN). We conducted a large-scale quantitative study from an existing coronary OCT dataset to demonstrate the superiority of our proposed framework over existing deep learning frameworks. In addition, we confirmed the generalizability of our framework by applying it to fish corneal images and rat retinal images, demonstrating its capability to super-resolve morphological details in eye imaging.

Low-Dose CT Image Synthesis for Domain Adaptation Imaging Using a Generative Adversarial Network With Noise Encoding Transfer Learning

Deep learning (DL) based image processing methods have been successfully applied to low-dose x-ray images based on the assumption that the feature distribution of the training data is consistent with that of the test data. However, low-dose computed tomography (LDCT) images from different commercial scanners may contain different amounts and types of image noise, violating this assumption. Moreover, in the application of DL based image processing methods to LDCT, the feature distributions of LDCT images from simulation and clinical CT examination can be quite different. Therefore, the network models trained with simulated image data or LDCT images from one specific scanner may not work well for another CT scanner and image processing task. To solve such domain adaptation problem, in this study, a novel generative adversarial network (GAN) with noise encoding transfer learning (NETL), or GAN-NETL, is proposed to generate a paired dataset with a different noise style. Specifically, we proposed a method to perform noise encoding operator and incorporate it into the generator to extract a noise style. Meanwhile, with a transfer learning (TL) approach, the image noise encoding operator transformed the noise type of the source domain to that of the target domain for realistic noise generation. One public and two private datasets are used to evaluate the proposed method. Experiment results demonstrated the feasibility and effectiveness of our proposed GAN-NETL model in LDCT image synthesis. In addition, we conduct additional image denoising study using the synthesized clinical LDCT data, which verified the merit of the proposed synthesis in improving the performance of the DL based LDCT processing method.

SNR-Net OCT: brighten and denoise low-light optical coherence tomography images via deep learning

Low-light optical coherence tomography (OCT) images generated when using low input power, low-quantum-efficiency detection units, low exposure time, or facing high-reflective surfaces, have low bright and signal-to-noise rates (SNR), and restrict OCT technique and clinical applications. While low input power, low quantum efficiency, and low exposure time can help reduce the hardware requirements and accelerate imaging speed; high-reflective surfaces are unavoidable sometimes. Here we propose a deep-learning-based technique to brighten and denoise low-light OCT images, termed SNR-Net OCT. The proposed SNR-Net OCT deeply integrated a conventional OCT setup and a residual-dense-block U-Net generative adversarial network with channel-wise attention connections trained using a customized large speckle-free SNR-enhanced brighter OCT dataset. Results demonstrated that the proposed SNR-Net OCT can brighten low-light OCT images and remove the speckle noise effectively, with enhancing SNR and maintaining the tissue microstructures well. Moreover, compared to the hardware-based techniques, the proposed SNR-Net OCT can be of lower cost and better performance.

Self-supervised Blind2Unblind deep learning scheme for OCT speckle reductions

As a low-coherence interferometry-based imaging modality, optical coherence tomography (OCT) inevitably suffers from the influence of speckles originating from multiply scattered photons. Speckles hide tissue microstructures and degrade the accuracy of disease diagnoses, which thus hinder OCT clinical applications. Various methods have been proposed to address such an issue, yet they suffer either from the heavy computational load, or the lack of high-quality clean images prior, or both. In this paper, a novel self-supervised deep learning scheme, namely, Blind2Unblind network with refinement strategy (B2Unet), is proposed for OCT speckle reduction with a single noisy image only. Specifically, the overall B2Unet network architecture is presented first, and then, a global-aware mask mapper together with a loss function are devised to improve image perception and optimize sampled mask mapper blind spots, respectively. To make the blind spots visible to B2Unet, a new re-visible loss is also designed, and its convergence is discussed with the speckle properties being considered. Extensive experiments with different OCT image datasets are finally conducted to compare B2Unet with those state-of-the-art existing methods. Both qualitative and quantitative results convincingly demonstrate that B2Unet outperforms the state-of-the-art model-based and fully supervised deep-learning methods, and it is robust and capable of effectively suppressing speckles while preserving the important tissue micro-structures in OCT images in different cases.

A self-supervised guided knowledge distillation framework for unpaired low-dose CT image denoising

Low-dose computed tomography (LDCT) can significantly reduce the damage of X-ray to the human body, but the reduction of CT dose will produce images with severe noise and artifacts, which will affect the diagnosis of doctors. Recently, deep learning has attracted more and more attention from researchers. However, most of the denoising networks applied to deep learning-based LDCT imaging are supervised methods, which require paired data for network training. In a realistic imaging scenario, obtaining well-aligned image pairs is challenging due to the error in the table re-positioning and the patient's physiological movement during data acquisition. In contrast, the unpaired learning method can overcome the drawbacks of supervised learning, making it more feasible to collect unpaired training data in most real-world imaging applications. In this study, we develop a novel unpaired learning framework, Self-Supervised Guided Knowledge Distillation (SGKD), which enables the guidance of supervised learning using the results generated by self-supervised learning. The proposed SGKD scheme contains two stages of network training. First, we can achieve the LDCT image quality improvement by the designed self-supervised cycle network. Meanwhile, it can also produce two complementary training datasets from the unpaired LDCT and NDCT images. Second, a knowledge distillation strategy with the above two datasets is exploited to further improve the LDCT image denoising performance. To evaluate the effectiveness and feasibility of the proposed method, extensive experiments were performed on the simulated AAPM challenging and real-world clinical LDCT datasets. The qualitative and quantitative results show that the proposed SGKD achieves better performance in terms of noise suppression and detail preservation compared with some state-of-the-art network models.

Self-supervised denoising of Nyquist-sampled volumetric images via deep learning

Purpose

Deep learning has demonstrated excellent performance enhancing noisy or degraded biomedical images. However, many of these models require access to a noise-free version of the images to provide supervision during training, which limits their utility. Here, we develop an algorithm (noise2Nyquist) that leverages the fact that Nyquist sampling provides guarantees about the maximum difference between adjacent slices in a volumetric image, which allows denoising to be performed without access to clean images. We aim to show that our method is more broadly applicable and more effective than other self-supervised denoising algorithms on real biomedical images, and provides comparable performance to algorithms that need clean images during training.

Approach

We first provide a theoretical analysis of noise2Nyquist and an upper bound for denoising error based on sampling rate. We go on to demonstrate its effectiveness in denoising in a simulated example as well as real fluorescence confocal microscopy, computed tomography, and optical coherence tomography images.

Results

We find that our method has better denoising performance than existing self-supervised methods and is applicable to datasets where clean versions are not available. Our method resulted in peak signal to noise ratio (PSNR) within 1 dB and structural similarity (SSIM) index within 0.02 of supervised methods. On medical images, it outperforms existing self-supervised methods by an average of 3 dB in PSNR and 0.1 in SSIM.

Conclusion

noise2Nyquist can be used to denoise any volumetric dataset sampled at at least the Nyquist rate making it useful for a wide variety of existing datasets.

Multi-task generative adversarial network for retinal optical coherence tomography image denoising

Objective. Optical coherence tomography (OCT) has become an essential imaging modality for the assessment of ophthalmic diseases. However, speckle noise in OCT images obscures subtle but important morphological details and hampers its clinical applications. In this work, a novel multi-task generative adversarial network (MGAN) is proposed for retinal OCT image denoising.Approach. To strengthen the preservation of retinal structural information in the OCT denoising procedure, the proposed MGAN integrates adversarial learning and multi-task learning. Specifically, the generator of MGAN simultaneously undertakes two tasks, including the denoising task and the segmentation task. The segmentation task aims at the generation of the retinal segmentation map, which can guide the denoising task to focus on the retina-related region based on the retina-attention module. In doing so, the denoising task can enhance the attention to the retinal region and subsequently protect the structural detail based on the supervision of the structural similarity index measure loss.Main results. The proposed MGAN was evaluated and analyzed on three public OCT datasets. The qualitative and quantitative comparisons show that the MGAN method can achieve higher image quality, and is more effective in both speckle noise reduction and structural information preservation than previous denoising methods.Significance. We have presented a MGAN for retinal OCT image denoising. The proposed method provides an effective way to strengthen the preservation of structural information while suppressing speckle noise, and can promote the OCT applications in the clinical observation and diagnosis of retinopathy.

Domain and Content Adaptive Convolution Based Multi-Source Domain Generalization for Medical Image Segmentation

The domain gap caused mainly by variable medical image quality renders a major obstacle on the path between training a segmentation model in the lab and applying the trained model to unseen clinical data. To address this issue, domain generalization methods have been proposed, which however usually use static convolutions and are less flexible. In this paper, we propose a multi-source domain generalization model based on the domain and content adaptive convolution (DCAC) for the segmentation of medical images across different modalities. Specifically, we design the domain adaptive convolution (DAC) module and content adaptive convolution (CAC) module and incorporate both into an encoder-decoder backbone. In the DAC module, a dynamic convolutional head is conditioned on the predicted domain code of the input to make our model adapt to the unseen target domain. In the CAC module, a dynamic convolutional head is conditioned on the global image features to make our model adapt to the test image. We evaluated the DCAC model against the baseline and four state-of-the-art domain generalization methods on the prostate segmentation, COVID-19 lesion segmentation, and optic cup/optic disc segmentation tasks. Our results not only indicate that the proposed DCAC model outperforms all competing methods on each segmentation task but also demonstrate the effectiveness of the DAC and CAC modules. Code is available at https://git.io/DCAC.

Self-supervised speckle noise reduction of optical coherence tomography without clean data

Optical coherence tomography (OCT) is widely used in clinical diagnosis due to its non-invasive, real-time, and high-resolution characteristics. However, the inherent speckle noise seriously degrades the image quality, which might damage the fine structures in OCT, thus affecting the diagnosis results. In recent years, supervised deep learning-based denoising methods have shown excellent denoising ability. To train a deep denoiser, a large number of paired noisy-clean images are required, which is difficult to achieve in clinical practice, since acquiring a speckle-free OCT image requires dozens of repeated scans and image registration. In this research, we propose a self-supervised strategy that helps build a despeckling model by training it to map neighboring pixels in a single noisy OCT image. Adjacent pixel patches are randomly selected from the original OCT image to generate two similar undersampled images, which are respectively used as the input and target images for training a deep neural network. To ensure both the despeckling and the structure-preserving effects, a multi-scale pixel patch sampler and corresponding loss functions are adopted in our practice. Through quantitative evaluation and qualitative visual comparison, we found that the proposed method performs better than state-of-the-art methods regarding despeckling effects and structure preservation. Besides, the proposed method is much easier to train and deploy without the need for clean OCT images, which has great significance in clinical practice.

Hybrid-structure network and network comparative study for deep-learning-based speckle-modulating optical coherence tomography

Optical coherence tomography (OCT), a promising noninvasive bioimaging technique, can resolve sample three-dimensional microstructures. However, speckle noise imposes obvious limitations on OCT resolving capabilities. Here we proposed a deep-learning-based speckle-modulating OCT based on a hybrid-structure network, residual-dense-block U-Net generative adversarial network (RDBU-Net GAN), and further conducted a comprehensively comparative study to explore multi-type deep-learning architectures' abilities to extract speckle pattern characteristics and remove speckle, and resolve microstructures. This is the first time that network comparative study has been performed on a customized dataset containing mass more-general speckle patterns obtained from a custom-built speckle-modulating OCT, but not on retinal OCT datasets with limited speckle patterns. Results demonstrated that the proposed RDBU-Net GAN has a more excellent ability to extract speckle pattern characteristics and remove speckle, and resolve microstructures. This work will be useful for future studies on OCT speckle removing and deep-learning-based speckle-modulating OCT.

A Novel Speckle Suppression Method with Quantitative Combination of Total Variation and Anisotropic Diffusion PDE Model

Speckle noise seriously affects synthetic aperture radar (SAR) image application. Speckle suppression aims to smooth the homogenous region while preserving edge and texture in the image. A novel speckle suppression method based on the combination of total variation and partial differential equation denoising models is proposed in this paper. Taking full account of the local statistics in the image, a quantization technique—which is different from the normal edge detection method—is supported by the variation coefficient of the image. Accordingly, a quantizer is designed to respond to both noise level and edge strength. This quantizer automatically determines the threshold of diffusion coefficient and controls the weight between total variation filter and anisotropic diffusion partial differential equation filter. A series of experiments are conducted to test the performance of the quantizer and proposed filter. Extensive experimental results have demonstrated the superiority of the proposed method with both synthetic images and natural SAR images.

Comparative study of deep neural networks with unsupervised Noise2Noise strategy for noise reduction of optical coherence tomography images

As a powerful diagnostic tool, optical coherence tomography (OCT) has been widely used in various clinical setting. However, OCT images are susceptible to inherent speckle noise that may contaminate subtle structure information, due to low-coherence interferometric imaging procedure. Many supervised learning-based models have achieved impressive performance in reducing speckle noise of OCT images trained with a large number of noisy-clean paired OCT images, which are not commonly feasible in clinical practice. In this article, we conducted a comparative study to investigate the denoising performance of OCT images over different deep neural networks through an unsupervised Noise2Noise (N2N) strategy, which only trained with noisy OCT samples. Four representative network architectures including U-shaped model, multi-information stream model, straight-information stream model and GAN-based model were investigated on an OCT image dataset acquired from healthy human eyes. The results demonstrated all four unsupervised N2N models offered denoised OCT images with a performance comparable with that of supervised learning models, illustrating the effectiveness of unsupervised N2N models in denoising OCT images. Furthermore, U-shaped models and GAN-based models using UNet network as generator are two preferred and suitable architectures for reducing speckle noise of OCT images and preserving fine structure information of retinal layers under unsupervised N2N circumstances.

Updates in deep learning research in ophthalmology

Ophthalmology has been one of the early adopters of artificial intelligence (AI) within the medical field. Deep learning (DL), in particular, has garnered significant attention due to the availability of large amounts of data and digitized ocular images. Currently, AI in Ophthalmology is mainly focused on improving disease classification and supporting decision-making when treating ophthalmic diseases such as diabetic retinopathy, age-related macular degeneration (AMD), glaucoma and retinopathy of prematurity (ROP). However, most of the DL systems (DLSs) developed thus far remain in the research stage and only a handful are able to achieve clinical translation. This phenomenon is due to a combination of factors including concerns over security and privacy, poor generalizability, trust and explainability issues, unfavorable end-user perceptions and uncertain economic value. Overcoming this challenge would require a combination approach. Firstly, emerging techniques such as federated learning (FL), generative adversarial networks (GANs), autonomous AI and blockchain will be playing an increasingly critical role to enhance privacy, collaboration and DLS performance. Next, compliance to reporting and regulatory guidelines, such as CONSORT-AI and STARD-AI, will be required to in order to improve transparency, minimize abuse and ensure reproducibility. Thirdly, frameworks will be required to obtain patient consent, perform ethical assessment and evaluate end-user perception. Lastly, proper health economic assessment (HEA) must be performed to provide financial visibility during the early phases of DLS development. This is necessary to manage resources prudently and guide the development of DLS.